2-Keto-L-gulonic acid which is useful as a precursor for the synthesis of vitamin C has been industrially produced by the Richstein method [Helvetica Chimica Acta 17, 311 (1934)]. However, this method involves many steps, and improvement in total yield can not be expected. Therefore, it is desired to find out more effective production methods.
Instead of the Reichstein method, a method which comprises producing 5-keto-L-gluconic acid from glucose microbiologically, converting it into idonic acid chemically or microbiologically, and then oxidizing the resultant to give 2-keto-L-gulonic acid (U.S. Pat. No. 2,421,611), and a method which comprises producing 2,5-diketo-D-gluconic acid from glucose microbiologically and reducing the resultant to 2-keto-L-gulonic acid chemically or microbiologically (Japanese Patent Publication Nos. 39-14493, 53-25033, 56-15877 and 59-35920) have been studied. However, the chemical reduction steps employed in these methods, are not stereospecific, and the former provides D-gluconic acid and the latter provides 2-keto-D-gluconic acid as by-products, which results in a decrease in yield. Further, when this step is carried out microbiologically, excessive carbon source should be supplied as a reduction energy source.
Also, the method for production of 2-keto-L-gulonic acid using L-sorbose as a starting material has been known. In this case, the product can be produced only by an oxidation step without involving any reduction step. As examples of this method, the methods using bacteria belonging to the genera Gluconobacter, Pseudomonas, Serratia, Achromobacter and Alcaligenes, have been heretofore known [see Biotechnology and Bioengineering 14, 799 (1972), Acta Microbiological Sinica, 20, 246 (1980) and 21, 185 (1981), Japanese Patent Publication Nos. 41-159 and 41-160, U.S. Pat. No. 3,043,749, Japanese Patent Publication No. 49-39838 and U.S.S.R. Patent No. 526,660]. However, the methods for production of 2-keto-L-gulonic acid from L-sorbose by strains which have been disclosed give extremely low product yield, and they can not be industrially utilized.
One improvement of these known methods by using a microorganism belonging to the genus Pseudogluconobacter is disclosed in the co-pending U.S. patent application Ser. No. 913,230 filed Oct. 1, 1986 which has been assigned to the same assignee as that of the present application.